1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
|
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2012, 2014 by Delphix. All rights reserved.
* Copyright (c) 2013 Steven Hartland. All rights reserved.
* Copyright (c) 2014 Integros [integros.com]
* Copyright 2017 RackTop Systems.
*/
/*
* LibZFS_Core (lzc) is intended to replace most functionality in libzfs.
* It has the following characteristics:
*
* - Thread Safe. libzfs_core is accessible concurrently from multiple
* threads. This is accomplished primarily by avoiding global data
* (e.g. caching). Since it's thread-safe, there is no reason for a
* process to have multiple libzfs "instances". Therefore, we store
* our few pieces of data (e.g. the file descriptor) in global
* variables. The fd is reference-counted so that the libzfs_core
* library can be "initialized" multiple times (e.g. by different
* consumers within the same process).
*
* - Committed Interface. The libzfs_core interface will be committed,
* therefore consumers can compile against it and be confident that
* their code will continue to work on future releases of this code.
* Currently, the interface is Evolving (not Committed), but we intend
* to commit to it once it is more complete and we determine that it
* meets the needs of all consumers.
*
* - Programatic Error Handling. libzfs_core communicates errors with
* defined error numbers, and doesn't print anything to stdout/stderr.
*
* - Thin Layer. libzfs_core is a thin layer, marshaling arguments
* to/from the kernel ioctls. There is generally a 1:1 correspondence
* between libzfs_core functions and ioctls to /dev/zfs.
*
* - Clear Atomicity. Because libzfs_core functions are generally 1:1
* with kernel ioctls, and kernel ioctls are general atomic, each
* libzfs_core function is atomic. For example, creating multiple
* snapshots with a single call to lzc_snapshot() is atomic -- it
* can't fail with only some of the requested snapshots created, even
* in the event of power loss or system crash.
*
* - Continued libzfs Support. Some higher-level operations (e.g.
* support for "zfs send -R") are too complicated to fit the scope of
* libzfs_core. This functionality will continue to live in libzfs.
* Where appropriate, libzfs will use the underlying atomic operations
* of libzfs_core. For example, libzfs may implement "zfs send -R |
* zfs receive" by using individual "send one snapshot", rename,
* destroy, and "receive one snapshot" operations in libzfs_core.
* /sbin/zfs and /zbin/zpool will link with both libzfs and
* libzfs_core. Other consumers should aim to use only libzfs_core,
* since that will be the supported, stable interface going forwards.
*/
#define _IN_LIBZFS_CORE_
#include <libzfs_core.h>
#include <ctype.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <fcntl.h>
#include <pthread.h>
#include <sys/nvpair.h>
#include <sys/param.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/zfs_ioctl.h>
#include "libzfs_core_compat.h"
#include "libzfs_compat.h"
#ifdef __FreeBSD__
extern int zfs_ioctl_version;
#endif
static int g_fd;
static pthread_mutex_t g_lock = PTHREAD_MUTEX_INITIALIZER;
static int g_refcount;
int
libzfs_core_init(void)
{
(void) pthread_mutex_lock(&g_lock);
if (g_refcount == 0) {
g_fd = open("/dev/zfs", O_RDWR);
if (g_fd < 0) {
(void) pthread_mutex_unlock(&g_lock);
return (errno);
}
}
g_refcount++;
(void) pthread_mutex_unlock(&g_lock);
return (0);
}
void
libzfs_core_fini(void)
{
(void) pthread_mutex_lock(&g_lock);
ASSERT3S(g_refcount, >, 0);
g_refcount--;
if (g_refcount == 0)
(void) close(g_fd);
(void) pthread_mutex_unlock(&g_lock);
}
static int
lzc_ioctl(zfs_ioc_t ioc, const char *name,
nvlist_t *source, nvlist_t **resultp)
{
zfs_cmd_t zc = { 0 };
int error = 0;
char *packed;
#ifdef __FreeBSD__
nvlist_t *oldsource;
#endif
size_t size;
ASSERT3S(g_refcount, >, 0);
(void) strlcpy(zc.zc_name, name, sizeof (zc.zc_name));
#ifdef __FreeBSD__
if (zfs_ioctl_version == ZFS_IOCVER_UNDEF)
zfs_ioctl_version = get_zfs_ioctl_version();
if (zfs_ioctl_version < ZFS_IOCVER_LZC) {
oldsource = source;
error = lzc_compat_pre(&zc, &ioc, &source);
if (error)
return (error);
}
#endif
packed = fnvlist_pack(source, &size);
zc.zc_nvlist_src = (uint64_t)(uintptr_t)packed;
zc.zc_nvlist_src_size = size;
if (resultp != NULL) {
*resultp = NULL;
zc.zc_nvlist_dst_size = MAX(size * 2, 128 * 1024);
zc.zc_nvlist_dst = (uint64_t)(uintptr_t)
malloc(zc.zc_nvlist_dst_size);
#ifdef illumos
if (zc.zc_nvlist_dst == NULL) {
#else
if (zc.zc_nvlist_dst == 0) {
#endif
error = ENOMEM;
goto out;
}
}
while (ioctl(g_fd, ioc, &zc) != 0) {
if (errno == ENOMEM && resultp != NULL) {
free((void *)(uintptr_t)zc.zc_nvlist_dst);
zc.zc_nvlist_dst_size *= 2;
zc.zc_nvlist_dst = (uint64_t)(uintptr_t)
malloc(zc.zc_nvlist_dst_size);
#ifdef illumos
if (zc.zc_nvlist_dst == NULL) {
#else
if (zc.zc_nvlist_dst == 0) {
#endif
error = ENOMEM;
goto out;
}
} else {
error = errno;
break;
}
}
#ifdef __FreeBSD__
if (zfs_ioctl_version < ZFS_IOCVER_LZC)
lzc_compat_post(&zc, ioc);
#endif
if (zc.zc_nvlist_dst_filled) {
*resultp = fnvlist_unpack((void *)(uintptr_t)zc.zc_nvlist_dst,
zc.zc_nvlist_dst_size);
}
#ifdef __FreeBSD__
if (zfs_ioctl_version < ZFS_IOCVER_LZC)
lzc_compat_outnvl(&zc, ioc, resultp);
#endif
out:
#ifdef __FreeBSD__
if (zfs_ioctl_version < ZFS_IOCVER_LZC) {
if (source != oldsource)
nvlist_free(source);
source = oldsource;
}
#endif
fnvlist_pack_free(packed, size);
free((void *)(uintptr_t)zc.zc_nvlist_dst);
return (error);
}
int
lzc_create(const char *fsname, enum lzc_dataset_type type, nvlist_t *props)
{
int error;
nvlist_t *args = fnvlist_alloc();
fnvlist_add_int32(args, "type", (dmu_objset_type_t)type);
if (props != NULL)
fnvlist_add_nvlist(args, "props", props);
error = lzc_ioctl(ZFS_IOC_CREATE, fsname, args, NULL);
nvlist_free(args);
return (error);
}
int
lzc_clone(const char *fsname, const char *origin,
nvlist_t *props)
{
int error;
nvlist_t *args = fnvlist_alloc();
fnvlist_add_string(args, "origin", origin);
if (props != NULL)
fnvlist_add_nvlist(args, "props", props);
error = lzc_ioctl(ZFS_IOC_CLONE, fsname, args, NULL);
nvlist_free(args);
return (error);
}
int
lzc_promote(const char *fsname, char *snapnamebuf, int snapnamelen)
{
/*
* The promote ioctl is still legacy, so we need to construct our
* own zfs_cmd_t rather than using lzc_ioctl().
*/
zfs_cmd_t zc = { 0 };
ASSERT3S(g_refcount, >, 0);
VERIFY3S(g_fd, !=, -1);
(void) strlcpy(zc.zc_name, fsname, sizeof (zc.zc_name));
if (ioctl(g_fd, ZFS_IOC_PROMOTE, &zc) != 0) {
int error = errno;
if (error == EEXIST && snapnamebuf != NULL)
(void) strlcpy(snapnamebuf, zc.zc_string, snapnamelen);
return (error);
}
return (0);
}
/*
* Creates snapshots.
*
* The keys in the snaps nvlist are the snapshots to be created.
* They must all be in the same pool.
*
* The props nvlist is properties to set. Currently only user properties
* are supported. { user:prop_name -> string value }
*
* The returned results nvlist will have an entry for each snapshot that failed.
* The value will be the (int32) error code.
*
* The return value will be 0 if all snapshots were created, otherwise it will
* be the errno of a (unspecified) snapshot that failed.
*/
int
lzc_snapshot(nvlist_t *snaps, nvlist_t *props, nvlist_t **errlist)
{
nvpair_t *elem;
nvlist_t *args;
int error;
char pool[ZFS_MAX_DATASET_NAME_LEN];
*errlist = NULL;
/* determine the pool name */
elem = nvlist_next_nvpair(snaps, NULL);
if (elem == NULL)
return (0);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/@")] = '\0';
args = fnvlist_alloc();
fnvlist_add_nvlist(args, "snaps", snaps);
if (props != NULL)
fnvlist_add_nvlist(args, "props", props);
error = lzc_ioctl(ZFS_IOC_SNAPSHOT, pool, args, errlist);
nvlist_free(args);
return (error);
}
/*
* Destroys snapshots.
*
* The keys in the snaps nvlist are the snapshots to be destroyed.
* They must all be in the same pool.
*
* Snapshots that do not exist will be silently ignored.
*
* If 'defer' is not set, and a snapshot has user holds or clones, the
* destroy operation will fail and none of the snapshots will be
* destroyed.
*
* If 'defer' is set, and a snapshot has user holds or clones, it will be
* marked for deferred destruction, and will be destroyed when the last hold
* or clone is removed/destroyed.
*
* The return value will be 0 if all snapshots were destroyed (or marked for
* later destruction if 'defer' is set) or didn't exist to begin with.
*
* Otherwise the return value will be the errno of a (unspecified) snapshot
* that failed, no snapshots will be destroyed, and the errlist will have an
* entry for each snapshot that failed. The value in the errlist will be
* the (int32) error code.
*/
int
lzc_destroy_snaps(nvlist_t *snaps, boolean_t defer, nvlist_t **errlist)
{
nvpair_t *elem;
nvlist_t *args;
int error;
char pool[ZFS_MAX_DATASET_NAME_LEN];
/* determine the pool name */
elem = nvlist_next_nvpair(snaps, NULL);
if (elem == NULL)
return (0);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/@")] = '\0';
args = fnvlist_alloc();
fnvlist_add_nvlist(args, "snaps", snaps);
if (defer)
fnvlist_add_boolean(args, "defer");
error = lzc_ioctl(ZFS_IOC_DESTROY_SNAPS, pool, args, errlist);
nvlist_free(args);
return (error);
}
int
lzc_snaprange_space(const char *firstsnap, const char *lastsnap,
uint64_t *usedp)
{
nvlist_t *args;
nvlist_t *result;
int err;
char fs[ZFS_MAX_DATASET_NAME_LEN];
char *atp;
/* determine the fs name */
(void) strlcpy(fs, firstsnap, sizeof (fs));
atp = strchr(fs, '@');
if (atp == NULL)
return (EINVAL);
*atp = '\0';
args = fnvlist_alloc();
fnvlist_add_string(args, "firstsnap", firstsnap);
err = lzc_ioctl(ZFS_IOC_SPACE_SNAPS, lastsnap, args, &result);
nvlist_free(args);
if (err == 0)
*usedp = fnvlist_lookup_uint64(result, "used");
fnvlist_free(result);
return (err);
}
boolean_t
lzc_exists(const char *dataset)
{
/*
* The objset_stats ioctl is still legacy, so we need to construct our
* own zfs_cmd_t rather than using lzc_ioctl().
*/
zfs_cmd_t zc = { 0 };
(void) strlcpy(zc.zc_name, dataset, sizeof (zc.zc_name));
return (ioctl(g_fd, ZFS_IOC_OBJSET_STATS, &zc) == 0);
}
/*
* Create "user holds" on snapshots. If there is a hold on a snapshot,
* the snapshot can not be destroyed. (However, it can be marked for deletion
* by lzc_destroy_snaps(defer=B_TRUE).)
*
* The keys in the nvlist are snapshot names.
* The snapshots must all be in the same pool.
* The value is the name of the hold (string type).
*
* If cleanup_fd is not -1, it must be the result of open("/dev/zfs", O_EXCL).
* In this case, when the cleanup_fd is closed (including on process
* termination), the holds will be released. If the system is shut down
* uncleanly, the holds will be released when the pool is next opened
* or imported.
*
* Holds for snapshots which don't exist will be skipped and have an entry
* added to errlist, but will not cause an overall failure.
*
* The return value will be 0 if all holds, for snapshots that existed,
* were succesfully created.
*
* Otherwise the return value will be the errno of a (unspecified) hold that
* failed and no holds will be created.
*
* In all cases the errlist will have an entry for each hold that failed
* (name = snapshot), with its value being the error code (int32).
*/
int
lzc_hold(nvlist_t *holds, int cleanup_fd, nvlist_t **errlist)
{
char pool[ZFS_MAX_DATASET_NAME_LEN];
nvlist_t *args;
nvpair_t *elem;
int error;
/* determine the pool name */
elem = nvlist_next_nvpair(holds, NULL);
if (elem == NULL)
return (0);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/@")] = '\0';
args = fnvlist_alloc();
fnvlist_add_nvlist(args, "holds", holds);
if (cleanup_fd != -1)
fnvlist_add_int32(args, "cleanup_fd", cleanup_fd);
error = lzc_ioctl(ZFS_IOC_HOLD, pool, args, errlist);
nvlist_free(args);
return (error);
}
/*
* Release "user holds" on snapshots. If the snapshot has been marked for
* deferred destroy (by lzc_destroy_snaps(defer=B_TRUE)), it does not have
* any clones, and all the user holds are removed, then the snapshot will be
* destroyed.
*
* The keys in the nvlist are snapshot names.
* The snapshots must all be in the same pool.
* The value is a nvlist whose keys are the holds to remove.
*
* Holds which failed to release because they didn't exist will have an entry
* added to errlist, but will not cause an overall failure.
*
* The return value will be 0 if the nvl holds was empty or all holds that
* existed, were successfully removed.
*
* Otherwise the return value will be the errno of a (unspecified) hold that
* failed to release and no holds will be released.
*
* In all cases the errlist will have an entry for each hold that failed to
* to release.
*/
int
lzc_release(nvlist_t *holds, nvlist_t **errlist)
{
char pool[ZFS_MAX_DATASET_NAME_LEN];
nvpair_t *elem;
/* determine the pool name */
elem = nvlist_next_nvpair(holds, NULL);
if (elem == NULL)
return (0);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/@")] = '\0';
return (lzc_ioctl(ZFS_IOC_RELEASE, pool, holds, errlist));
}
/*
* Retrieve list of user holds on the specified snapshot.
*
* On success, *holdsp will be set to a nvlist which the caller must free.
* The keys are the names of the holds, and the value is the creation time
* of the hold (uint64) in seconds since the epoch.
*/
int
lzc_get_holds(const char *snapname, nvlist_t **holdsp)
{
int error;
nvlist_t *innvl = fnvlist_alloc();
error = lzc_ioctl(ZFS_IOC_GET_HOLDS, snapname, innvl, holdsp);
fnvlist_free(innvl);
return (error);
}
/*
* Generate a zfs send stream for the specified snapshot and write it to
* the specified file descriptor.
*
* "snapname" is the full name of the snapshot to send (e.g. "pool/fs@snap")
*
* If "from" is NULL, a full (non-incremental) stream will be sent.
* If "from" is non-NULL, it must be the full name of a snapshot or
* bookmark to send an incremental from (e.g. "pool/fs@earlier_snap" or
* "pool/fs#earlier_bmark"). If non-NULL, the specified snapshot or
* bookmark must represent an earlier point in the history of "snapname").
* It can be an earlier snapshot in the same filesystem or zvol as "snapname",
* or it can be the origin of "snapname"'s filesystem, or an earlier
* snapshot in the origin, etc.
*
* "fd" is the file descriptor to write the send stream to.
*
* If "flags" contains LZC_SEND_FLAG_LARGE_BLOCK, the stream is permitted
* to contain DRR_WRITE records with drr_length > 128K, and DRR_OBJECT
* records with drr_blksz > 128K.
*
* If "flags" contains LZC_SEND_FLAG_EMBED_DATA, the stream is permitted
* to contain DRR_WRITE_EMBEDDED records with drr_etype==BP_EMBEDDED_TYPE_DATA,
* which the receiving system must support (as indicated by support
* for the "embedded_data" feature).
*/
int
lzc_send(const char *snapname, const char *from, int fd,
enum lzc_send_flags flags)
{
return (lzc_send_resume(snapname, from, fd, flags, 0, 0));
}
int
lzc_send_resume(const char *snapname, const char *from, int fd,
enum lzc_send_flags flags, uint64_t resumeobj, uint64_t resumeoff)
{
nvlist_t *args;
int err;
args = fnvlist_alloc();
fnvlist_add_int32(args, "fd", fd);
if (from != NULL)
fnvlist_add_string(args, "fromsnap", from);
if (flags & LZC_SEND_FLAG_LARGE_BLOCK)
fnvlist_add_boolean(args, "largeblockok");
if (flags & LZC_SEND_FLAG_EMBED_DATA)
fnvlist_add_boolean(args, "embedok");
if (flags & LZC_SEND_FLAG_COMPRESS)
fnvlist_add_boolean(args, "compressok");
if (resumeobj != 0 || resumeoff != 0) {
fnvlist_add_uint64(args, "resume_object", resumeobj);
fnvlist_add_uint64(args, "resume_offset", resumeoff);
}
err = lzc_ioctl(ZFS_IOC_SEND_NEW, snapname, args, NULL);
nvlist_free(args);
return (err);
}
/*
* "from" can be NULL, a snapshot, or a bookmark.
*
* If from is NULL, a full (non-incremental) stream will be estimated. This
* is calculated very efficiently.
*
* If from is a snapshot, lzc_send_space uses the deadlists attached to
* each snapshot to efficiently estimate the stream size.
*
* If from is a bookmark, the indirect blocks in the destination snapshot
* are traversed, looking for blocks with a birth time since the creation TXG of
* the snapshot this bookmark was created from. This will result in
* significantly more I/O and be less efficient than a send space estimation on
* an equivalent snapshot.
*/
int
lzc_send_space(const char *snapname, const char *from,
enum lzc_send_flags flags, uint64_t *spacep)
{
nvlist_t *args;
nvlist_t *result;
int err;
args = fnvlist_alloc();
if (from != NULL)
fnvlist_add_string(args, "from", from);
if (flags & LZC_SEND_FLAG_LARGE_BLOCK)
fnvlist_add_boolean(args, "largeblockok");
if (flags & LZC_SEND_FLAG_EMBED_DATA)
fnvlist_add_boolean(args, "embedok");
if (flags & LZC_SEND_FLAG_COMPRESS)
fnvlist_add_boolean(args, "compressok");
err = lzc_ioctl(ZFS_IOC_SEND_SPACE, snapname, args, &result);
nvlist_free(args);
if (err == 0)
*spacep = fnvlist_lookup_uint64(result, "space");
nvlist_free(result);
return (err);
}
static int
recv_read(int fd, void *buf, int ilen)
{
char *cp = buf;
int rv;
int len = ilen;
do {
rv = read(fd, cp, len);
cp += rv;
len -= rv;
} while (rv > 0);
if (rv < 0 || len != 0)
return (EIO);
return (0);
}
static int
recv_impl(const char *snapname, nvlist_t *props, const char *origin,
boolean_t force, boolean_t resumable, int fd,
const dmu_replay_record_t *begin_record)
{
/*
* The receive ioctl is still legacy, so we need to construct our own
* zfs_cmd_t rather than using zfsc_ioctl().
*/
zfs_cmd_t zc = { 0 };
char *atp;
char *packed = NULL;
size_t size;
int error;
ASSERT3S(g_refcount, >, 0);
/* zc_name is name of containing filesystem */
(void) strlcpy(zc.zc_name, snapname, sizeof (zc.zc_name));
atp = strchr(zc.zc_name, '@');
if (atp == NULL)
return (EINVAL);
*atp = '\0';
/* if the fs does not exist, try its parent. */
if (!lzc_exists(zc.zc_name)) {
char *slashp = strrchr(zc.zc_name, '/');
if (slashp == NULL)
return (ENOENT);
*slashp = '\0';
}
/* zc_value is full name of the snapshot to create */
(void) strlcpy(zc.zc_value, snapname, sizeof (zc.zc_value));
if (props != NULL) {
/* zc_nvlist_src is props to set */
packed = fnvlist_pack(props, &size);
zc.zc_nvlist_src = (uint64_t)(uintptr_t)packed;
zc.zc_nvlist_src_size = size;
}
/* zc_string is name of clone origin (if DRR_FLAG_CLONE) */
if (origin != NULL)
(void) strlcpy(zc.zc_string, origin, sizeof (zc.zc_string));
/* zc_begin_record is non-byteswapped BEGIN record */
if (begin_record == NULL) {
error = recv_read(fd, &zc.zc_begin_record,
sizeof (zc.zc_begin_record));
if (error != 0)
goto out;
} else {
zc.zc_begin_record = *begin_record;
}
/* zc_cookie is fd to read from */
zc.zc_cookie = fd;
/* zc guid is force flag */
zc.zc_guid = force;
zc.zc_resumable = resumable;
/* zc_cleanup_fd is unused */
zc.zc_cleanup_fd = -1;
error = ioctl(g_fd, ZFS_IOC_RECV, &zc);
if (error != 0)
error = errno;
out:
if (packed != NULL)
fnvlist_pack_free(packed, size);
free((void*)(uintptr_t)zc.zc_nvlist_dst);
return (error);
}
/*
* The simplest receive case: receive from the specified fd, creating the
* specified snapshot. Apply the specified properties as "received" properties
* (which can be overridden by locally-set properties). If the stream is a
* clone, its origin snapshot must be specified by 'origin'. The 'force'
* flag will cause the target filesystem to be rolled back or destroyed if
* necessary to receive.
*
* Return 0 on success or an errno on failure.
*
* Note: this interface does not work on dedup'd streams
* (those with DMU_BACKUP_FEATURE_DEDUP).
*/
int
lzc_receive(const char *snapname, nvlist_t *props, const char *origin,
boolean_t force, int fd)
{
return (recv_impl(snapname, props, origin, force, B_FALSE, fd, NULL));
}
/*
* Like lzc_receive, but if the receive fails due to premature stream
* termination, the intermediate state will be preserved on disk. In this
* case, ECKSUM will be returned. The receive may subsequently be resumed
* with a resuming send stream generated by lzc_send_resume().
*/
int
lzc_receive_resumable(const char *snapname, nvlist_t *props, const char *origin,
boolean_t force, int fd)
{
return (recv_impl(snapname, props, origin, force, B_TRUE, fd, NULL));
}
/*
* Like lzc_receive, but allows the caller to read the begin record and then to
* pass it in. That could be useful if the caller wants to derive, for example,
* the snapname or the origin parameters based on the information contained in
* the begin record.
* The begin record must be in its original form as read from the stream,
* in other words, it should not be byteswapped.
*
* The 'resumable' parameter allows to obtain the same behavior as with
* lzc_receive_resumable.
*/
int
lzc_receive_with_header(const char *snapname, nvlist_t *props,
const char *origin, boolean_t force, boolean_t resumable, int fd,
const dmu_replay_record_t *begin_record)
{
if (begin_record == NULL)
return (EINVAL);
return (recv_impl(snapname, props, origin, force, resumable, fd,
begin_record));
}
/*
* Roll back this filesystem or volume to its most recent snapshot.
* If snapnamebuf is not NULL, it will be filled in with the name
* of the most recent snapshot.
* Note that the latest snapshot may change if a new one is concurrently
* created or the current one is destroyed. lzc_rollback_to can be used
* to roll back to a specific latest snapshot.
*
* Return 0 on success or an errno on failure.
*/
int
lzc_rollback(const char *fsname, char *snapnamebuf, int snapnamelen)
{
nvlist_t *args;
nvlist_t *result;
int err;
args = fnvlist_alloc();
err = lzc_ioctl(ZFS_IOC_ROLLBACK, fsname, args, &result);
nvlist_free(args);
if (err == 0 && snapnamebuf != NULL) {
const char *snapname = fnvlist_lookup_string(result, "target");
(void) strlcpy(snapnamebuf, snapname, snapnamelen);
}
nvlist_free(result);
return (err);
}
/*
* Roll back this filesystem or volume to the specified snapshot,
* if possible.
*
* Return 0 on success or an errno on failure.
*/
int
lzc_rollback_to(const char *fsname, const char *snapname)
{
nvlist_t *args;
nvlist_t *result;
int err;
args = fnvlist_alloc();
fnvlist_add_string(args, "target", snapname);
err = lzc_ioctl(ZFS_IOC_ROLLBACK, fsname, args, &result);
nvlist_free(args);
nvlist_free(result);
return (err);
}
/*
* Creates bookmarks.
*
* The bookmarks nvlist maps from name of the bookmark (e.g. "pool/fs#bmark") to
* the name of the snapshot (e.g. "pool/fs@snap"). All the bookmarks and
* snapshots must be in the same pool.
*
* The returned results nvlist will have an entry for each bookmark that failed.
* The value will be the (int32) error code.
*
* The return value will be 0 if all bookmarks were created, otherwise it will
* be the errno of a (undetermined) bookmarks that failed.
*/
int
lzc_bookmark(nvlist_t *bookmarks, nvlist_t **errlist)
{
nvpair_t *elem;
int error;
char pool[ZFS_MAX_DATASET_NAME_LEN];
/* determine the pool name */
elem = nvlist_next_nvpair(bookmarks, NULL);
if (elem == NULL)
return (0);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/#")] = '\0';
error = lzc_ioctl(ZFS_IOC_BOOKMARK, pool, bookmarks, errlist);
return (error);
}
/*
* Retrieve bookmarks.
*
* Retrieve the list of bookmarks for the given file system. The props
* parameter is an nvlist of property names (with no values) that will be
* returned for each bookmark.
*
* The following are valid properties on bookmarks, all of which are numbers
* (represented as uint64 in the nvlist)
*
* "guid" - globally unique identifier of the snapshot it refers to
* "createtxg" - txg when the snapshot it refers to was created
* "creation" - timestamp when the snapshot it refers to was created
*
* The format of the returned nvlist as follows:
* <short name of bookmark> -> {
* <name of property> -> {
* "value" -> uint64
* }
* }
*/
int
lzc_get_bookmarks(const char *fsname, nvlist_t *props, nvlist_t **bmarks)
{
return (lzc_ioctl(ZFS_IOC_GET_BOOKMARKS, fsname, props, bmarks));
}
/*
* Destroys bookmarks.
*
* The keys in the bmarks nvlist are the bookmarks to be destroyed.
* They must all be in the same pool. Bookmarks are specified as
* <fs>#<bmark>.
*
* Bookmarks that do not exist will be silently ignored.
*
* The return value will be 0 if all bookmarks that existed were destroyed.
*
* Otherwise the return value will be the errno of a (undetermined) bookmark
* that failed, no bookmarks will be destroyed, and the errlist will have an
* entry for each bookmarks that failed. The value in the errlist will be
* the (int32) error code.
*/
int
lzc_destroy_bookmarks(nvlist_t *bmarks, nvlist_t **errlist)
{
nvpair_t *elem;
int error;
char pool[ZFS_MAX_DATASET_NAME_LEN];
/* determine the pool name */
elem = nvlist_next_nvpair(bmarks, NULL);
if (elem == NULL)
return (0);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/#")] = '\0';
error = lzc_ioctl(ZFS_IOC_DESTROY_BOOKMARKS, pool, bmarks, errlist);
return (error);
}
|